Bacterial community distribution and functional potentials provide key insights into their role in the ecosystem functioning of a retreating Eastern Himalayan glacier

Abstract Himalayan glaciers are receding at an exceptional rate, perturbing the local biome and ecosystem processes. Understanding the microbial ecology of an exclusively microbe-driven biome provides insights into their contributions to the ecosystem functioning through biogeochemical fluxes. Here, we investigated the bacterial communities and their functional potential in the retreating East Rathong Glacier (ERG) of Sikkim Himalaya. Amplicon-based taxonomic classification revealed the dominance of the phyla Proteobacteria, Bacteroidota, and candidate Patescibacteria in the glacial sites. Further, eight good-quality metagenome-assembled genomes (MAGs) of Proteobacteria, Patescibacteria, Acidobacteriota, and Choloflexota retrieved from the metagenomes elucidated the microbial contributions to nutrient cycling. The ERG MAGs showed aerobic respiration as a primary metabolic feature, accompanied by carbon fixation and complex carbon degradation potentials. Pathways for nitrogen metabolism, chiefly dissimilatory nitrate reduction and denitrification, and a complete sulphur oxidation enzyme complex for sulphur metabolism were identified in the MAGs. We observed that DNA repair and oxidative stress response genes complemented with osmotic and periplasmic stress and protein chaperones were vital for adaptation against the intense radiation and stress conditions of the extreme Himalayan niche. Current findings elucidate the microbiome and associated functional potentials of a vulnerable glacier, emphasizing their significant ecological roles in a changing glacial ecosystem.


Introduction
Ever since the past century, a constant rise in global temperature has resulted in climate change (Jansson and Hofmockel 2020 ).Reports estimate that global warming may exceed the le v el of 1.5 • C or 2 • C by the next decade (IPCC 2022) .T he Himala yan regions ar e particularl y mor e vulner able to climate c hange than the global mean (Bajr ac harya and Shr estha 2011 ).One major consequence of climate change is the shrinkage of glaciers, which dir ectl y impacts the downstream ecosystems as mountain glaciers ar e essential r egulators of the global atmospheric, hydr ological, and biogeochemical cycles (Milner et al. 2017 ).Studies on the impact of deglaciation have exclusively been focussed on macrole v el dynamics, while r esearc h concerning its effect on glacial micr obial comm unities is still at its nascency.The study of the glacier microbiome is significant as glaciers are recognized as unique biomes dri ven exclusi vely by microorganisms through global biogeochemical fluxes (Anesio andLaybourn-Parry 2012 , Anesio et al. 2017 ).Micr obial div ersity holds a dynamic r elationship with ecosystem functioning as they play multifunctional roles in pri-mary production, nutrient cycling, organic decomposition, and climate regulation (Delgado-Baquerizo et al. 2016 ).Dominant micr obial comm unities of the supr a glacial surface ar e k e y dri vers of the production and transport of dissolved organic carbon and nitr ogen fr om the glacier to the downstream ecosystems for sustaining food webs (Milner et al. 2017 ).These micr oor ganisms sequester the atmospheric labile inorganic nutrients, metabolizing them into an organic carbon pool and forming a hotspot for primary production (Anesio et al. 2009 ).Further, microbial colonizers are the pioneers of primary succession in deglaciated barren forefields , pa ving the wa y for plant and other complex community colonization (Ciccazzo et al. 2016 ).The expediting global temperature rise is accompanied by changes in precipitation and glacier cov er a ge that ar e likel y to alter the native microbial community and their functional roles in primary productivity and nutrient cycling in the glacier system (Rathore et al. 2022 ).These facts indicate that any loss of microbial diversity can lead to inefficient functioning of the terrestrial as well as aquatic ecosystems (Bodelier 2011 , Cardinale et al. 2011 ).
While ecological and functional information on microbial life is widely covered from the glaciers and permafrost of Arctic and Antarctic regions (Edwards et al. 2013, Lutz et al. 2017, Li et al. 2019, Xue et al. 2020, Wu et al. 2021 ), Indian Himalayan glaciers ar e r elativ el y less inv estigated.Some studies hav e been conducted in the glaciers of Indian Western Himalaya to gain insights into the microbial distribution patterns and their potential functional roles (Kumar et al. 2019, 2022, Bhattacharya et al. 2022, Rathore et al. 2022 ).Recent micr obiome anal ysis of Changme Khang and Changme Khangpu in the Sikkim Himalaya has r e v ealed a rich micr obial div ersity with scope for future biotechnological potential (Sherpa et al. 2019(Sherpa et al. , 2020 ) ). Considering the significance of microbial diversity and the ecological roles they perform, the microbial community dynamics of the supraglacial and forefield regions of Sikkim Himalayan glaciers are rarely explored.A dominant reason for this lack of study is the remote and obscure location of the Himala yan glaciers , accompanied by rugged and r ough hill y terr ains (Rathor e et al. 2022 ).The curr ent r esearc h is conducted on one such high-altitude glacier in Sikkim, the East Rathong Glacier (ERG)-a benchmark glacier in the Eastern Himalaya for long-term monitoring of glacier mass balance, and hydrological balance owing to factors like ideal length, size, accessibility, and visible impact of climate change .T he glacier retreated by 460 m within a timespan of 1980-2012 at an av er a ge r ate of 13.3 m/year (Luitel et al. 2012 ).The present study aims to assess the microbial community composition and their functional potentials from the ablation zone supr a glacial surface and the closest forefield site at an ele v ation gr adient of 4600-4700 masl.We attempt to ac hie v e a holistic view of the metabolic and str ess ada ptation str ategies of the microbial communities, apprehending their significance in ecosystem functioning.

Study site and field sampling
The ERG is a south-east facing valley type glacier located between 88 • 06 27.63 E and 27 • 34 54.44 N within the Kanchenjunga National Park of Sikkim in the Eastern Himalaya (Mukhia et al. 2022a, Sharma et al. 2022 ).Covering an area of 4.80 km 2 and ∼7 km in length, the glacier is debris-free and a source of the river Rangit.It takes a 39-km trek to reach the study site from the last accessible village, Yuksom, in West Sikkim.
Sampling was conducted in pre-October 2021, encompassing the supr a glacial and pr oglacial sites at an ele v ation of > 4600 masl.We designated the sampling sites as: Supr a glacial site-ERG1 and ERG2, 4670 masl; Proglacial site-ERG3 and ERG4, 4648-4657 masl (Table 1 and Fig. 1 ).Specificall y, pr oglacial sediment samples were obtained from bare and dry areas, whereas supr a glacial samples were systematically collected along run-off c hannels.At the Supr a glacial site, w e collected ice-meltw ater and sediment samples from two points (ERG1 and ERG2) on the glacier surface of the ablation zone.At the Proglacial site, soil sediments were collected from two points (ERG3 and ERG4) across the barren moraine in the immediate vicinity of the glacier snout.At each point, sediment sample collection was done in triplicates within a distance of ∼1 m, while the water sample was collected in duplicates .T he ice meltwater sample (depth of 10-15 cm) was collected in a Sterivex filter unit (Millipore, USA), and the same was stored in 50 ml tubes (Tarsons, India) for physicochemical analyses, while sediment samples (depth of 5-10 cm) were collected in sterile sample bags .T he samples wer e tr ansited with ice pac ks till Yuksom and further to the laboratory within 48 h by air.

Physicochemical analyses of samples
The pH and electrical conductivity (EC) of samples were determined using a digital pH and EC meter (Eutech PC 450, Thermo Scientific , USA).T he sediment pH and EC wer e measur ed by preparing suspensions of 1:2.5 (w/v) (Broadbent et al. 2021 ) and 1:5 (w/v) (Bockheim 2007 ) in deionized water, r espectiv el y.Sediment samples were oven-dried, sieved, and cleared of rocks and gravel for c hemical anal yses .T he total carbon (TC), nitrogen (TN), and sulphur (S) contents of the solid samples wer e measur ed using the CHNS elemental analyzer (Vario MICRO cube, Elementar, Germany).The elemental composition of the solid and liquid samples was determined using an Atomic Absor ption Spectr ometer (Shimadzu AA-6300, Japan) after the extraction process following the hot aqua regia digestion method (Rice et al. 2017, Kumari et al. 2022 ).

DN A extr action
Total DNA from the sediment and water samples was extracted using a FastDNA spin kit for soil (MP biomedicals, California, USA) according to the manufacturer's guidelines with few modifications like bead-beating at maximum speed for 10 min and prolonged pr otein pr ecipitation at 4 • C to incr ease the yield.Extr acted DN A w as subjected to nanodrop (Thermofisher Scientific, USA) and a gar ose gel assessments befor e PCR amplification.

Amplicon and shotgun sequencing
For microbiome analysis, amplicon sequencing of supraglacial and proglacial samples was performed for the V3-V4 region of the 16S rRNA gene using the primer pair V13F: -5 A GA GTTTGAT-GMTGGCTCAG 3 and V13R: -5 TTA CCGCGGCMGCSGGCA C 3 (Saxena et al. 2021 ).The PCR conditions included: initial denaturation at 95 • C, 25 cycles of denaturation at 95 • C for 15 s, annealing at 60 • C for 15 s, elongation at 72 • C for 2 min, and a final extension at 72 • C for 10 min, using 12.5 ng template DNA.The amplicons fr om eac h sample wer e purified with AMPur e XP beads to r emov e unused primers.Sequencing libr aries wer e pr epar ed using the Nextera XT DNA Library Preparation kit (Illumina, USA), following the manufacturer's instructions.Libraries were purified using AMPure XP beads and quantified using a Qubit dsDNA highsensitivity assay kit.Sequencing was performed on the Illumina MiSeq platform (Illumina).
To analyze the functional potential of the microbiota, shotgun sequencing was performed on two soil sediment samples (one eac h fr om the supr a glacial and pr oglacial sites).Meta genome libr aries wer e pr epar ed using the KAP A HyperPlus kit (KAP A Biosystems, USA), following the manufacturer's instructions with 400 ng template DNA.The DNA was fr a gmented using the KAPA fr a gmentation method into 600 bp length fr a gments .T he fr a gmented samples wer e pr ocessed for end r epair and A-tailing with the Hy-perPrep/HyperPlus ERAT enzyme mix.Immediately after the end repair and A-tailing, adapter ligation to the end-repaired DNA fr a gments using DNA ligase was performed.Postligation cleanup was performed using 0.7X AMPure XP beads to r emov e an y unincor por ated ada pter.Libr ary amplification of the ada pter-ligated DN A samples w as done using Illumina primers.Sequencing was performed using the Illumina Hiseq 4000 platform (Illumina).

Data processing and taxonomic assignments
T he ra w reads generated from amplicon sequencing were demultiplexed to r emov e the barcodes using an in-house script.Quality c hec k of the r eads was done using FastQC v0.11.9 ( https:// www.bioinformatics.babraham.ac.uk/pr ojects/fastqc/).Onl y forw ar d reads were considered due to the suboptimal quality of the Table 1.Description of sampling sites and sample types at each site along with physicochemical parameters (average ± standard deviation) of each sample; masl: metres above sea level, TC: total carbon, TN: total nitrogen, NA: not applicable/not measured.r e v erse r eads .T he adapters , primers , and sequence reads with a phred score < 30 were removed using Cutadapt v3.4 (Martin 2011 ).Further trimmed reads were analyzed using the QIIME2 v2021.2pipeline (Bolyen et al. 2019 ).The denoising and c himer a r emov al fr om the imported single-end r eads wer e ac hie v ed using the D AD A2 pipeline (Callahan et al. 2016 ) ( Table S1 , Supporting Information ).For taxonomic assignments, we used the SILVA v138 database (McDonald et al. 2012 ) with q2-feature-classifier (Bokulich et al. 2018 ).The feature table and feature sequences wer e filter ed using 'qiime taxa filter -table' and 'qiime taxa filterseqs' commands for removing unassigned amplicon sequence variants (ASVs) and those annotated as mitochondria.The prefilter ed, r ar efied ASVs table and taxonomy were used for the calculation of Alpha diversity indices (Shannon index, Simpson index, and species observed) using the Phyloseq R pac ka ge v1.46.0 (McMurdie and Holmes 2013 ).Additionally, the effect of different environmental factors on the microbial abundance of ERG was e v aluated by redundancy analysis (RDA).The RDA was performed b y emplo ying micr oeco R pac ka ge v1.2.0, whic h uses the rda function of the v egan pac ka ge (Oksanen et al. 2019 ).To test the significant differences in the bacterial diversity between the sample groups, a Kruskal-Wallis Rank Sum test was performed at a 95% significance le v el, using the function Kruskal test in R v4.2.The w orkflo w is provided in Table S2 ( Supporting Information ).

Metagenome binning and analysis
T he ra w metagenomic reads were processed using the KBase server (Arkin et al. 2018 ) according to a pr e viousl y described method (Chivian et al. 2022 ).The r aw r eads wer e coassembled using the metaSPAdes v3.15.3 tool (Nurk et al. 2017 ), and binning of the metagenome contigs was achieved by using Maxbin2 v2.2.4 (Wu et al. 2016 ), MetaBAT2 v1.7 (Kang et al. 2019 ), and CONCOCT v1.1 (Alneberg et al. 2013 ).The quality of the bins generated from the three tools was improved by employing the DAS Tool (Sieber et al. 2018 ) and assessed by the CheckM v1.0.18 tool (Parks et al. 2015 ).The bins with good or medium quality (completeness ≥ 50%, contamination < 10%) (Bo w ers et al. 2017 ) were retrieved and taxonomically classified using GTDB-Tk v1.7.0 (Chaumeil et al. 2020 ).A phylogenetic tree of the MAGs was prepared using a pr e viousl y described method (Wang et al. 2023 ).Briefly, the universal marker genes in the MAGs were identified using the identity module of GTDB-Tk (v2.3.2) and aligned using the align module of GTDB-Tk.FastTree (v2.1) was employed to construct the phylogenetic tree of the MAGs using the concatenated universal gene alignment under the WAG + GAMMA model (Price et al. 2010 ).Next, the phylogenetic tree was imported into iTOL (Letunic and Bork 2016 ) for additional r efinements.Furthermor e, to assess the functional potential of the bins in the glacier ecosystem, the bins wer e pr ocessed using the METABOLIC v4.0 tool (Zhou et al. 2022 ).Ad ditionally, the n ucleotide sequences of the eight MAGs were uploaded to the RAST server to get insights into the adaptation potential of the bins using subsystem technology (Aziz et al. 2008, Overbeek et al. 2014, Brettin et al. 2015 ).The w orkflo w is provided in Table S2 ( Supporting Information ).

Physicochemical profile of glacial samples
The samples sho w ed v ariations in physicoc hemical par ameters across the glacial sites (Table 1 ).The pH of the samples was mostly consistent in the range of 6.16-6.58, the farther proglacial sample being the most acidic, while the conductivity increased linearly along the supr a glacial to proglacial sites from 6.14 to 12.56 μS/cm.The analysis of solid sediment samples sho w ed lo w TC (0.11%-0.19%) and S (0.05%-0.07%) and undetectable TN contents.Further elemental anal ysis r e v ealed the absence of elements like F e , K, Mn, and Zn and low concentrations of other elements ranging from 0.06 ppm (Cu) to 12.17 ppm (Ca) in the supr a glacial meltwater sample .T he elements F e and K were the most abundant, follo w ed b y Mg, Mn, Ca, Zn, Cu, and P in the sediment samples of both the supr a glacial and proglacial sites.No traces of heavy metals like Pb, Ni, Cr, or Cd were detected in any of the samples, depicting the pristine and uncontaminated nature of the sites.

Bacterial community composition and diversity in the ERG
A total of 501 085 good-quality single-end sequence reads with a length of 300 nucleotides were obtained.Post denoising and c himer a r emov al, 350 588 r eads wer e obtained, whic h r esulted in 622 ASVs.In contr ast, pair ed-end sequence analysis yielded a substantially lo w er initial sequence reads of 49 735 resulting in 32 777 reads and 128 ASVs after denoising and c himer a r emov al. Ther efore , to a void the risk of data loss , forw ar d-onl y r eads wer e considered.
The alpha diversity indices (Shannon, Simpson, and Observed) indicated the highest diversity in the farther proglacial sample (ERG4) and the supr a glacial meltwater sample (ERG1), with the observed species being highly represented in ERG4.The supr a glacial sediments (ERG2) and the near-to-snout proglacial samples (ERG3) sho w ed similar r epr esentations of the div ersity indices (Fig. 2 C).The Kruskal-Wallis rank sum test suggested no significant difference in the bacterial alpha diversity between the sample groups at 95% confidence le v el.Further, RDA was a pplied

Metagenome-assembled genome reconstruction from the ERG metagenomes
We selected two sites, namel y, the supr a glacial sample of the ablation zone-ERG2 and the proglacial sample farther from the glacier snout-ERG4, for metagenome sequencing and analysis.The ERG2 and ERG4 samples have 26 182 518 and 34 271 933 raw r eads, r espectiv el y.After quality trimming of the reads, the ERG2 sample was left with 20 359 146 (77.8%) good-quality reads, while sample ERG4 was left with 27 226 791 (79.4%) good-quality reads.The metaSPAdes assembly consists of 21 489 contigs from both samples, with the largest contig size of 639 764 bp and N50 of 7746.A total of one high-quality (completeness > 90% and contamination < 10%) and se v en medium-quality (completeness > 50% and contamination < 10%) metagenome-assembled genomes (MAGs) wer e r etrie v ed fr om the supr a glacial and pr oglacial meta genomes that were further analyzed.Among the eight selected MAGs, four sho w ed affiliations with the phylum Proteobacteria (bin05, bin08, bin10, and bin12), two with Patescibacteria (bin04 and bin06), and one each with Acidobacteriota (bin03) and Chloroflexota (bin11) as r e v ealed by GTDB-Tk analysis (Table 2 ).The genome statistics are provided in Table S3 ( Supporting Information ).The size of MAGs ranged from 0.8 to 5.4 Mb, with a GC content ranging from 46% to 68%.The Proteobacterial MAGs were further resolved into four gener a, namel y Massilia , Sphingomicrobium , Rhizobacter , and Novosphingobium (Fig. 3 ).The Patescibacterial MAGs were as-sociated with lineages within the order Saccharimonadales (genera classified as RGVC01 and UBA4729).The remaining MAGs of Acidobacteriota and Chloroflexota corresponded to classes Thermoanaerobaculia and Ellin6529 and are classified as uncultivated genus Fen-183 and family CSP1-4, respectively.

Metabolic functions of MAGs regulating the nutrient cycles in ERG
The metabolic pathways of the MAGs wer e decipher ed using the METABOLIC tool ( Table S4 , Supporting Information ).The tool enables the prediction of the metabolic and biogeochemical functional profiles of the bins.We found that the ERG MAGs are activ el y involv ed in the cycling of n utrients, as de picted by the KEGG module step hits (Fig. 4 A and B).Aerobic respiration was predominant in all the MAGs, as evidenced by the occurrence of genes for gl ycol ysis , tricarboxylic acid cycle , pyruvate oxidation, and o xidati ve phosphorylation.Ho w ever, Patescibacterial MAGs wer e r estricted to gl ycol ysis and o xidati ve phosphorylation.The ability of MAGs to fix CO 2 into organic molecules was depicted by the presence of genes for five carbon fixation pathwa ys , i.e .Calvin-Benson-Bassham (CBB), Wood-Ljungdahl, r eductiv e tricarbo xylic acid (rTCA), 3-hydro xypropionate, and dicarbo xylatehydr oxybutyr ate .While the CBB pathwa y w as w ell r epr esented in all eight MAGs, Patescibacterial MAGs did not carry genes for the other four pathwa ys .T he Wood-Ljungdahl pathwa y that fixes CO 2 or CO or other C1 carbon into acetyl-CoA was only exhibited by the Proteobacterial MAGs .T he Proteobacterial MAGs' potential to metabolize C1 molecules as a carbon source was reflected by the genes for methane oxidation ( mxaF/mdh1 and mxaI/mdh2 ).Additionall y, Pr oteobacterial MAGs ( Massilia , Rhizobacter and Novosphingobium ) sho w ed the potential for complex carbon degradation, as reflected by the degradation genes for pectin, T he pathwa ys associated with nitrogen cycling were detected in all the MAGs assigned to Proteobacteria, Acidobacteriota, and Chlor oflexota, while completel y absent in Patescibacteria (Fig. 4 B).The MAGs r etrie v ed fr om ERG meta genomes wer e limited to dissimilatory and assimilatory nitr ate r eduction and denitrificationrelated pathwa ys .Notably, Acidobacteriota and Proteobacteria ( Massilia ) contained most nitrogen metabolism genes.A complete denitrification potential was observed in these MAGs by encoding periplasmic nitr ate r eductase Na pAB, nitr ate r eductase NarGHI, nitrite reductase NirK and NirS, nitric oxide reductase NorBC, CYP55, and nitrous oxide reductase NosZ.Similarly, the two MAGs sho w ed full potential for dissimilatory nitrate reduction as they contained nitr ate r eductase Na pAB, NarGHI, and nitrite reductase NirBD, NrfAH.Only Massilia and Rhizobacter carried partial genes for assimilatory nitr ate r eduction to ammonia (NarB, NR, and NasAB).None of the MAGs sho w ed the capacity for nitrogen fixation.
The sulphur cycling pathway was absent in Patescibacterial MA G , while it was highl y r epr esented in the Proteobacterial MAG of Massilia (Fig. 4 B).Chiefly, Massilia sho w ed the capacity for complete assimilatory sulfate reduction (CysND, Sat, PAPSS , CysC , CysH, CysJI, and Sir), while other Proteobacteria and Chloroflexota MAGs sho w ed partial assimilatory and dissimilatory sulfate reduction (CysND and Sat) potential.Moreover, a complete sulphur oxidation (SOX) enzyme complex (SoxA, SoxX, SoxB, SoxC, So xY, and So xZ) was observ ed in the Acidobacteriota and Pr oteobacterial MAGs .T he SOX complex is capable of oxidizing thiosulfate , sulfite , sulfide , or elemental sulphur to sulfate (Wang et al. 2019 ).

Adaptation of ERG MAGs to environmental stress
RAST analysis of the MAGs for proteins involved in stress response r e v ealed their distribution into categories like DNA repair, oxidativ e str ess, osmotic str ess, periplasmic str ess, pr otein c ha per ones, and carbohydr ate starv ation ( Table S5 , Supporting Information ).For DNA r epair a gainst UV or desiccation, genes encoding nucleotide exc hange r epair pr oteins , i.e .UvrABC excinuclease , were identified in the MAGs except for Sphingomicrobium and Novosphingobium .Mismatc h r epair pr oteins MutL-MutS and nonhomologous end-joining repair genes ( ligC , ligD , and ku domain protein) were detected in Acidobacteriota, Patescibacteria (g_UBA4729), and Proteobacteria ( Massilia , Sphingomicrobium , and Novosphingobium ).Additionall y, homologous r epair genes ( recFOR ) were present in Patescibacteria (g_RGVC01) and Proteobacteria MAGs.Many o xidati v e str ess r esponse pr oteins, including super oxide dism utase and ruberyththrin/rubr edoxin, wer e pr esent in a majority of the MAGs .Besides , glutathione reductase and glutathione peroxidase for the redox cycle and glutathione S-transferase and lacto ylglutathione ly ase for nonr edox r eactions wer e identified in the Proteobacterial MAGs.In response to osmotic stress, the MAGs of Acidobacteriota, Patescibacteria (g_UBA4729), and Proteobacteria ( Massilia , Sphingomicrobium , and Rhizobacter ) were equipped with cyclic beta-1,2-glucan synthase for the synthesis of Acidobacteriota, Chloroflexota, P atescibacteria, and Pr oteobacteria.The tr ee was inferr ed based on the concatenated gene alignment of univ ersal marker genes using the GTDB-Tk (v2.3.2).FastTree (v2.1) was employed to construct the phylogenetic tree, and iTOL was used for the final r epr esentation.osmoregulated periplasmic glucans .Besides , choline and betaine uptake and betaine biosynthesis proteins, including sarcosine oxidase, choline dehydrogenase, and choline uptake protein BetT, were detected in Novosphingobium .Proteins for combating periplasmic str ess wer e also identified, chiefly the HtrA protease, which has both chaperone and proteolytic activities for remo ving misfolded proteins .Others , like outer membr ane str ess sensor protease DegS and outer membrane protein H precursor, wer e pr esent in Pr oteobacteria ( Massilia and Rhizobacter ).Excluding Chloroflexota, all MAGs possessed the k e y chaperone proteins DnaJ and DnaK.Mor eov er, the genomes of Acidobacteriota and Patescibacteria carried the genes for carbon starvation protein A, which is known to be a pyruvate transporter (Gasperotti et al. 2020 ).

Discussion
Micr oor ganisms ar e an integr al part of the biospher e that driv es the ecological processes through the global biogeochemical cycling of nutrients.Glaciers and ice sheets, as cryosphere components , alone co ver ∼10% of the Earth's surface (IPCC 2019) .Being extr emel y sensitiv e to the warming climate, the acceler ated r etreat of alpine glaciers and ice masses would undoubtedly perturb the local biome and the associated ecosystem processes.Against this bac kdr op, the cryospheric micr obiome is of m uc h significance yet remains one of the most cryptic and poorly characterized among other microbiomes (Bourquin et al. 2022 ).To further our understanding of the alpine microbial communities and their roles in ecosystem functioning, our study investigates the transect along an ablation zone supr a glacial and nearby proglacial site of a vulnerable Himalayan glacier.
A glacier ecosystem includes supr a glacial and proglacial habitats with varying physicochemical properties.Our microbiome analysis of samples on the ablation zone of ERG r e v ealed man y shar ed bacterial phyla, v arying onl y in r elativ e abundance .T he dominant phyla, Proteobacteria and Bacteroidota, observed in the supr a glacial ice meltwater samples ar e pr edominant in the ice samples collected from other glaciers as well (Wilhelm et al. 2013, Lutz et al. 2015, Garcia-Lopez et al. 2019 ).The pr e v alent Patescibacteria in supraglacial and proglacial sediment samples constitute an unc har acterized gr oup fr equentl y documented in cold environments (Kumar et al. 2022, Rathore et al. 2022 ).Cyanobacteria, identified as the main primary producers in supr a glacial habitats worldwide (Anesio et al. 2017, Rathore et al. 2022, Jaarsma et al. 2023 ), was detected in the ERG surface sediments.Inter estingl y, Bdellovibrionota-a pr edatory bacterial gr oup, occur ed in the supr a glacial meltwater samples.Pr e viousl y, this bacteria has been r eported fr om Antarctic soils, marine waters, and perialpine lakes (Paix et al. 2019, Li et al. 2021, Ortiz et al. 2021 ).The predominance of the genus Pseudomonas in the supr a glacial meltwater corr obor ates our pr e vious cultur ede pendent stud y (Mukhia et al. 2022b ).The uncultured gen us Saccharimonadales LWQ8 was more enriched in the sediment samples of supr a glacial and pr oglacial sites.As anticipated, the pr oglacial site farther from the snout sho w ed the highest bacterial alpha diversity as it is a compar ativ el y older soil to hav e de v eloped after the glacial r etr eat featuring gr eater micr obial activities for bio w eathering.The supr a glacial meltw ater sample sho w ed the next highest alpha diversity.The surface ice is subject to the invasion of minerals and microbial cells from the wind, while the icemeltw ater provides w ater for micr obial gr owth and metabolism that together facilitates microbial colonization in this zone (Ane-sio and Laybourn-Parry 2012 ).Ecologists often use the concept of the source-sink hypothesis to estimate the flow of micr oor ganisms between habitats that might explain the shaping of observed differences (Burns et al. 2016 ).Source-sink dynamics could be pivotal in shaping the observed community dynamics within a studied ecosystem (Ezzat et al. 2022, Rolli et al. 2022 ).In this context, supr a glacial habitats may serve as diversity sources, suppl ying micr oor ganisms to pr oglacial sink habitats.Subsequentl y, the differ ential envir onmental conditions , nutrient a vailability, and physical c har acteristics betw een sour ce and sink habitats facilitate the unique establishments, impacting the entire ecosystem dynamics.To unveil the inclusive intricacies of the glacial bacterial comm unities, we delv ed into the meta genomes of r epr esentativ e supr a glacial and pr oglacial samples .T he observed phyla of the reconstructed ERG MAGs, namely Proteobacteria, P atescibacteria, Acidobacteriota, and Chlor oflexota, ar e often r ecov er ed fr om other glaciers and permafrosts as well (Xue et al. 2019, 2020, Varliero et al. 2021, Busi et al. 2022 ).Affirming our findings of the amplicon data, we r etrie v ed two MAGs belonging to the order Saccharimonadales and one each of the genus Massilia and Rhizobacter .Fascinatingly, the MAG within Acidobacteriota was annotated to the class Thermoanaer obaculia, whic h consists of thermophilic bacteria pr e viousl y isolated exclusiv el y fr om thermal habitats (Dedysh and La wson 2020 ).T he phylum Chloroflexota was observed in both amplicon and metagenome samples and the r etrie v ed MAG was taxonomicall y labeled up to the famil y le v el as CSP1-4.The sequences in this clade mostly originate from soil or river sediments as per SILVA taxonomy (Mehrshad et al. 2018 ).
The geogr a phical location of the high-altitude Himalayan glaciers r epr esents a hostile envir onment for life to flourish.How micr oor ganisms ada pt to suc h extr eme conditions and perform metabolic acti vities dri ving the ecosystem is indeed intriguing.
As glacier ecology is primaril y driv en by micr obial metabolism, elucidation of metabolic pathways is k e y to demonstrating the role of bacterial communities in niche-specific biogeochemical cycles .Our metagenomic in vestigation of MAGs focused on unr av eling the k e y metabolic strategies adopted to meet energy and carbon r equir ements.To no sur prise, all the identified MAGs sho w ed gr eat ca pacity for aer obic r espir ation, whic h highlights the pr eferred mode of energy fulfillment in the ERG communities.None of the MAGs demonstrated any noticeable ability for fermentation.Such a finding was in concurrence with other reports from Arctic and Antarctic soils (Xue et al. 2020, Ortiz et al. 2021 ).Physiological analysis revealed a shallow organic carbon content of 0.11%-0.19% in the ablation zone of the ERG r egion.Concurr entl y, we observed all the major carbon fixation pathways in the recov er ed MAGs, particularl y in Pr oteobacteria.Consequentl y, carbon dioxide fixation seems to be the dominant carbon uptake and metabolism mode instead of organic carbon oxidation in the nutrient-limited glacial en vironment.T he pr e v alence of div erse CO 2 -fixation pathways in the glacial MAGs pr ovides e vidence for efficient autotr ophic potential.Mainl y, thr ee MAGs ( Massilia , Sphingomicrobium , and Rhizobacter ) within Proteobacteria sho w ed an almost complete CBB cycle, including the central enzyme Ru-BisCO, which suggests the dominance of this pathway among all.Another glacier ecosystem has gained a similar inter pr etation (Trivedi et al. 2020 ).The next common carbon fixation pathway in the ERG may be the r eductiv e TCA cycle, as the genes for the pathw ay w er e also nearl y complete in Acidobacteriota and Proteobacteria ( Rhizobacter and Novosphingobium ) MAGs .T he other carbon metabolism pathways in the MAGs were mostly partial and may be secondary processes for carbon incorporation.
As inter pr eted fr om the r esults, dissimilatory nitr ate r eduction and denitrification may be the dominant nitrogen metabolism pathways in the r ecov er ed MAGs.The findings were similar to those observed in other Arctic glacier MAGs (Trivedi et al. 2020, Tian et al. 2022 ).While dissimilatory nitrate reduction to ammonium is a means of energy conservation that retains the N in the system for crucial biological processes, denitrification releases N back into the atmosphere.Both functions are essential for maintaining the nitrogen dynamics in an ecosystem.Most denitrifiers are known to switch from aerobic to anaerobic respiration under O 2 -limited conditions .T he glacier surface and forefield at the ablation zone are subject to changes in aerobic and anaerobic conditions .T his occurs due to intense environmental fluctuations consisting of heavy snow and rainfall follo w ed b y dry periods of vigorous light intensity (Tian et al. 2022 ).The complete dissimilatory nitr ate-r educing and denitrifying genes in some of the MAGs suggest their potential for an anaerobic mode of respiration to better adapt to the changing environmental conditions.Denitrification is active in many other alpine glaciers (Chen et al. 2022, Murakami et al. 2022 ).Ho w e v er, N fixation was not detected in the MAGs, as was observed in the Canadian High Arctic permafrost and Svalbard forefield MAGs (Wu et al. 2021, Tian et al. 2022 ), which might suggest the minimal contribution of the process in the nitrogen c ycle.Although, the FixL-FixJ tw o-component regulatory system was identified in the MAGs of Proteobacteria ( Massilia and Rhizobacter ) that stimulate nitrogen fixation under low oxygen conditions.
Sulphur metabolism pathways in the MAGs were dominated by assimilatory sulfate reduction and SOX, as suggested by the presence of complete pathway genes.Assimilatory sulfate reduction results in the biosynthesis of sulphur-containing amino acids in anoxygenic bacteria.The snow melting during summers enhances the weathering of r oc ks that contribute to sulfate content, whic h is activ el y metabolized by the c hemolithotr ophs (Rathor e et al. 2022 ).Hence, ERG MAGs are better involved in the utilization of inorganic sulphur for biosynthesis pr ocesses.Like wise, SOX is a vital aspect of biogeochemical sulphur cycling, where thiosulfate oxidation leads to the formation of elemental sulphur or sulfate.We found that the Acidobacteriota and Proteobacteria MAGs sho w ed the full potential for this means of energy gain.Gamma pr oteobacteria MAGs hav e been r eported for similar activity in a Canadian High Arctic permafrost (Wu et al. 2021 ).Some of the MAGs carried partial genes for dissimilatory sulfate reduction, which indicates the rarity of this pathway but might aid in the bio w eathering and release of nutrients for sustaining the sulphur cycle.Among all MAGs, Patescibacteria contained the least metabolism pathways that may be attributed to the obligately symbiotic lifestyle of the phylum and, hence, is deficient in alternative pathways for carbon acquisition or energy conservation (Ortiz et al. 2021 ).
To get insights into the basis of microbial life in ERG, we checked the stress response genes in the supraglacial and proglacial MAGs.Besides low-temper atur e conditions, the micr oor ganisms m ust surviv e thr ough intense UV irr adiations and o xidati v e str ess, desiccation, lo w w ater activity, osmotic stress, and low nutrient availability in high-altitude glaciers (Collins and Margesin 2019 ).We observed that the most pr e v alent genes wer e associated with DNA repair and oxidative stress response.DNA repair genes are crucial for microbial survival against UV and desiccation stress (Liu et al. 2022 ).Consequently, we detected uvrABC , mutS , mutL , lexA , ligC , and ligD , and other r epair pr oteins like r adA , r adC , recA , recN , recF , recO , and recR .Low envir onmental temper atur e is often linked with high oxidative stress (Chattopadhyay et al. 2011 ).Most MAGs carried superoxide dismutase, which is a k e y antio xidant enzyme for o xidati v e str ess toler ance in aer obic or ganisms.The enrichment of glutathione S-transferase, known to quench r eactiv e molecules, signifies the efficient regulation of oxidative stress in glacier bacteria, protecting the cells from o xidati ve burst and harmful effects of UV radiation (Kumar and Trivedi 2018 ).For osmotic str ess ada ptation, cyclic beta-1,2-glucan synthase particularl y occurr ed in most MAGs .T his enzyme is responsible for synthesizing a group of osmoregulated periplasmic glucans involved in osmoadaptation (Bontemps-Gallo et al. 2017 ).We found that the Proteobacteria MAGs are well-equipped with periplasmic str ess sensor pr oteins like degS and ompH since the microbial cell envelope is the first barrier and primary defense against environmental assaults .T he molecular c ha per one pr oteins dnaJ and dnaK that facilitate bacterial survival in a stressful environment (Mukhia et al. 2022a ) occurred uniformly in the glacial MAGs.

Conclusions
With r a pidl y c hanging ecosystems, the micr obial dynamics of the supr a glacial sites and immediate proglacial sites need to be understood as the glacial r etr eat influences the nutrient fluxes at the catchment sites, which later develop into a new ecosystem.To our knowledge, this is the first bacterial metagenomic assessment of the ERG.The glacial sites sho w ed major dominance of bacterial phyla Pr oteobacteria, Bacter oidetes, and P atescibacteria.The observation of di verse adapti ve and metabolic traits in the r ecov er ed MAGs demonstr ates the persistence of the resident micr oflor a for successful colonization in the oligotrophic stressful en vironment.T he presence of multiple carbon fixation pathways indicates the significance of c hemolithoautotr ophic Pr oteobacteria in the ecosystem.Aerobic respiration and nitrogen and sulphur metabolism potential were prevalent in the MAGs .T his signifies their metabolic flexibility for perse v er ance and ener gy conserv ation, contributing collectiv el y to the ecosystem functioning.Altogether, the detected nutrient c ycling pathw ays add to our knowledge of the roles of bacteria in global geochemical fluxes.

Figure 1 .
Figure 1.Location of the study area.(A) Map of sampling sites across the transect in the ERG, where ice-meltwater and sediment samples were collected in duplicates and triplicates, r espectiv el y, fr om eac h site.(B) Supr a glacial site and, (C) Pr oglacial site .T he ma p was gener ated in QGIS v3.26.2.

Figure 2 .
Figure 2. Taxonomic classification and diversity of bacterial communities across the supraglacial and proglacial samples of ERG.(A) Bacterial taxonomic composition at the phylum le v el.(B) Bacterial taxonomic composition at the genus le v el.(C) Alpha diversity indices (Observed, Shannon, and Simpson) showing the highest bacterial diversity in the farther proglacial site.(D) RDA plot indicating the correlation between the abundant bacterial genera and selective sample physicochemical parameters.

Figure 3 .
Figure 3. Phylogenetic tree of the eight MAGs recovered from the ERG showing affiliation to four phyla, viz.Acidobacteriota, Chloroflexota, P atescibacteria, and Pr oteobacteria.The tr ee was inferr ed based on the concatenated gene alignment of univ ersal marker genes using the GTDB-Tk (v2.3.2).FastTree (v2.1) was employed to construct the phylogenetic tree, and iTOL was used for the final r epr esentation.

Figure 4 .
Figure 4. Functional analysis of eight recovered MAGs using METABOLIC v4.0 tool.(A) Heatmap representation of the selected metabolic functions in each MAG according to the KEGG module, where the dark-blue box indicates a complete or nearly complete pathway and the light-blue box indicates an incomplete or absent pathway.A nearly complete pathway indicates the presence of most of the steps in the metabolic pathway, as inferred from the detection of annotated proteins.Role of different bacterial MAGs or taxa in (B) carbon, nitrogen, and sulphur cycling in the glacial ecosystem based on the presence and absence of genes given by KEGG module step hit.The coloured boxes represent individual MAGs, red-coloured line indicates the pr esence, while blac k-colour ed line indicates an absence or an incomplete pathway.

Table 2 .
Statistics of reconstructed MAGs determined using GTDB-tk.